BabyVision

A multimodal benchmark for fine-grained visual perception and grounded reasoning tasks.

8 models

Higher is better.

1Qwen3.8 MaxAlibaba82.0
2Muse Spark OpenMeta80.2
3Muse Spark 1.1Meta76.3
4Seed 2.1 ProByteDance73.7
5Qwen3.8-27BAlibaba65.7
6GLM-5.3-FlashZ.AI53.4
7dots3-note PreviewDots Studio50.0
8Nemotron 3.5 Super VL 120B A12BNVIDIA29.6

Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric: Multimodal visual reasoning
Tasks: Visual perception tasks
Authors: Meta AI (2026)
Source: Muse Spark 1.1 Evaluation Report ↗

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.